The treatment of a number of diseases including hemoglobinopathies and malaria will require a fundamental understanding of human erythropoiesis. Many experimental methodologies aimed at understanding this process are inherently flawed as they involve nonhuman cells or cell lines derived from transformed cells. The study of erythropoiesis in primary erythroblasts has been handicapped by experimental pitfalls associated with the ex vivo culture of those cells. We have taken a direct approach toward the prospective study of the early transcriptional events that encompass human erythropoiesis. Using flow cytometry to analyze liquid cultured blood from normal volunteers, we have identified and temporally phenotyped the erythroid continuum of cells present in these mass cultures. This approach has led to the identification of erythroblasts that are transcriptionally committed to erythroid differentiation. Surprisingly, within that population are cells able to form giant colonies in semisolid media suggesting differentiation and proliferation are not invariably coupled in normal human hematopoiesis. The most immature hematopoietic that have committed themselves toward the transcription of erythroid-specific genes represent a pivitol cell for the study of proliferation and differentiation events associated with normal and abnormal human erythropoiesis. Our immediate future goal is the characterization of these cells by antigenic and further transcriptional phenotyping. Once the transcriptional events associated with normal erythropoiesis are defined, correlates aimed at understanding disease states involving red blood cells may be possible.